LAUSR

Abstract As an important part of sustainable society, smart campuses have attracted more and more attention. By associating artificial intelligence with school infrastructure, it makes teaching services more convenient and personalized. Smart classroom is one of the important solutions of smart campuses because teaching and learning are always the primary tasks of smart campuses. At present, many intelligent technologies have been put into smart classrooms for monitoring and analyzing teaching activities, but the complexity of classroom behavior still impedes the development of smart classrooms. Existing researches on learning behavior analytics only focus on student actions. However, student learning behaviors are often closely related to teacher behaviors, so it is not objective and accurate to measure engagement only by analyzing student actions. In this paper, we propose a teacher-student behavioral engagement pattern (TSBEP) to synthetically measure student engagement by adding teacher behaviors. And the decision tree model based on classification and regression tree (CART) is used to predict engagement levels referring to the TSBEP. Experimental results show that the proposed TSBEP can measure the teaching and learning more accurately. And it is helpful to sustainably deploy and implement learning analytics to improve the quality of learning and teaching in smart classrooms.